House Structure Is Associated with Plasmodium Falciparum Infection in a Low-Transmission Setting in Southern Zambia

Messiah University Mosaic

Biology Educator Scholarship Biological Sciences

2016

House Structure is Associated With Plasmodium Falciparum Infection in a Low-Transmission Setting in Southern Zambia

M. M. Ippolito

K. M. Searle

H. Hamapumbu

T. M. Shields

J. C. Stevenson

See next page for additional authors

Follow this and additional works at: https://mosaic.messiah.edu/bio_ed

Part of the Biology Commons Permanent URL: https://mosaic.messiah.edu/bio_ed/112

Recommended Citation Ippolito, M. M.; Searle, K. M.; Hamapumbu, H.; Shields, T. M.; Stevenson, J. C.; Thuma, Philip; and Moss, W. J., "House Structure is Associated With Plasmodium Falciparum Infection in a Low-Transmission Setting in Southern Zambia" (2016). Biology Educator Scholarship. 112. https://mosaic.messiah.edu/bio_ed/112

Sharpening Intellect | Deepening Christian Faith | Inspiring Action

Messiah University is a Christian university of the liberal and applied arts and sciences. Our mission is to educate men and women toward maturity of intellect, character and Christian faith in preparation for lives of service, leadership and reconciliation in church and society. www.Messiah.edu One University Ave. | Mechanicsburg PA 17055 Authors M. M. Ippolito, K. M. Searle, H. Hamapumbu, T. M. Shields, J. C. Stevenson, Philip Thuma, and W. J. Moss

This article is available at Mosaic: https://mosaic.messiah.edu/bio_ed/112 Am. J. Trop. Med. Hyg., 97(5), 2017, pp. 1561–1567 doi:10.4269/ajtmh.17-0299 Copyright © 2017 by The American Society of Tropical Medicine and Hygiene

House Structure Is Associated with Plasmodium falciparum Infection in a Low-Transmission Setting in Southern Zambia

Matthew M. Ippolito,1* Kelly M. Searle,2 Harry Hamapumbu,3 Timothy M. Shields,2 Jennifer C. Stevenson,3,4 Philip E. Thuma,3 and William J. Moss2; for the Southern Africa International Center of Excellence for Malaria Research 1Divisions of Infectious Diseases and Clinical Pharmacology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland; 2Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland; 3Macha Research Trust, Macha Hospital, Choma District, Zambia; 4The W. Harry Feinstone Department of Molecular Microbiology and Immunology, The Johns Hopkins Malaria Research Institute, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland

Abstract. House structure may influence the risk of malaria by affecting mosquito entry and indoor resting. Identifi- cation of construction features associated with protective benefits could inform vector control approaches, even in low- transmission settings. We examined the association between house structure and malaria prevalence in a cross-sectional analysis of 2,788 children and adults residing in 866 houses in a low-transmission area of Southern Province, Zambia, over the period 2008–2012. Houses were categorized according to wall (brick/cement block or mud/grass) and roof (metal or grass) material. Malaria was assessed by point-of-care rapid diagnostic test (RDT) for Plasmodium falciparum.We identified 52 RDT-positive individuals residing in 41 houses, indicating an overall prevalence in the sample of 1.9%, ranging from 1.4% to 8.8% among the different house types. Occupants of higher quality houses had reduced odds of P. falciparum malaria compared with those in the lowest quality houses after controlling for bed net use, indoor insecticide spraying, clustering by house, cohabitation with another RDT-positive individual, transmission season, ecologic risk defined as nearest distance to a Strahler-classified third-order stream, education, age, and gender (adjusted odds ratio [OR]: 0.26, 95% confidence interval [CI]: 0.09–0.73, P = 0.01 for houses with brick/cement block walls and metal roof; OR: 0.22, 95% CI: 0.09–0.52, P < 0.01 for houses with brick/cement block walls and grass roof). Housing improvements offer a promising approach to vector control in low-transmission settings that circumvents the threat posed by insecticide resistance, and may confer a protective benefit of similar magnitude to current vector control strategies.

INTRODUCTION window screening or other entry barriers compared with those that did not.11–14 Malaria remains an important cause of morbidity and mor- However, studies in sub-Saharan Africa that examined as- tality in endemic regions worldwide, and vector control strat- 1,2 sociations of wall and roof construction with malaria have egies are vital to control and elimination efforts. The two yielded equivocal results. About half of the studies demon- predominantly deployed vector control measures are indoor strated an association, among which wall material appeared residual spraying (IRS) of insecticides and insecticide-treated fl 15–38 3 to be more in uential than roof material. Results of bed nets (ITNs). The emergence of insecticide resistance and studies that applied adjusted models to account for age, changes in behavior of mosquitoes to avoid contact with in- gender, ITN use, ecologic variables, and socioeconomic insecticides may threaten the efficacy of IRS and ITNs, creating 4 dicators were somewhat more conclusive; most demon- appeal for additional approaches to prevent malaria. strated a significant protective effect of high-quality walls Malaria is transmitted by female anopheline mosquito ranging from 24% to 63% reduction in the risk or odds of vectors that generally prefer to feed in the late evening and malaria, and half showed a protective effect of high-quality night and exhibit endophagic (indoor feeding) behavior, roofs ranging from 15% to 62% reduction.27–37 making the house a potentially high-risk transmission envi- 5 Results of a cross-sectional analysis of housing, grouped ronment. Housing features that impede mosquito entry and by wall and roof type, and malaria in a low-transmission area of indoor mosquito resting are, therefore, likely to diminish oc- southern Zambia are presented. Survey data and field ob- ’ 6,7 cupants risk of malaria. Indeed, housing improvements servations were analyzed from participants living in various such as window and door screening played an important role house types to inform potential approaches to housing inter- fi in malaria control programs during the rst half of the twentieth ventions for vector control against malaria in Zambia and century in North America and Europe before the widespread similar low-transmission settings in sub-Saharan Africa. 8 fi use of insecticides. The rst such experiments were con- Higher quality housing was hypothesized to correlate with ducted over a century ago by Angelo Celli in Italy, who rec- reduced prevalence of malaria compared with lower quality fi ognized malaria as a disease of poverty and identi ed poor housing. housing as a modifiable risk factor.9,10 More recently, his results were recapitulated in a small number of trials done in sub- Saharan Africa where malaria remains endemic, showing METHODS reduced numbers of indoor anopheline mosquitos and lower prevalence of childhood anemia in houses that received Study site. The study was conducted in a 1,200 km2 region east of Macha Hospital in Choma District, Southern Province, Zambia. The area lies at an altitude of 1,000 m above sea level * Address correspondence to Matthew M. Ippolito, Divisions of and the local biome is mainly Miombo woodland. The rainy Infectious Diseases and Clinical Pharmacology, Department of Medicine, Johns Hopkins University School of Medicine, 1830 E. season is from November to April, followed by a cool, dry Monument Street, Room. 450-B, Baltimore, MD 21231. E-mail: season from April to August and a hot, dry season from August [email protected] to November.39 The inhabitants are traditional villagers living 1561 1562 IPPOLITO AND OTHERS in homesteads consisting of one or more houses where a sampling frame for the random selection of homesteads. members of a family or extended family reside. In general, the Images were imported into ArcGIS 9.2 (Redlands, CA) and houses in the study area have doors or other makeshift bar- homestead locations were identified, manually enumerated, riers. Windows, when present, rarely have glass or screens but and randomly selected from the sampling frame for some windows have curtains. Eaves, a gap between the roof assignment to either the cross-sectional survey or the longi- and top edge of the wall, are open in nearly all houses con- tudinal cohort. The field team was provided with maps and structed with grass roofs, whereas most houses with metal Global Positioning System coordinates of the randomly roofs have closed eaves. selected homesteads. Transmission intensity in the study area is low. During the For each study visit, permission was obtained from the head study period, the entomological inoculation rate was < 1 in- of household, individual residents of the homestead were fective bite per person per season.40 The predominant malaria enumerated, and written informed consent was obtained from vector is Anopheles arabiensis.40 Vector control efforts in- each adult participant, or from the participant’s parent or clude distribution of ITNs, with little IRS having been carried guardian for children £ 18 years. Surveys were administered to out in the Macha area. Malaria control efforts include case gather individual-level demographic information and ITN use, management with artemisinin-based combination therapy, and house- and homestead-level information including prior introduced in Zambia in 2002 and into the study area in application of IRS, educational achievement of the head of the 2004.41,42 household, and availability of flush toilet and electricity. ITN Study design. The study was conducted within the context use was determined by an affirmative answer to the survey of an epidemiologic survey of malaria using data collected item, “Do you sleep under a bed net?” Homestead distance to from February 2008 to February 2012.43 Homesteads in the Strahler-classified third-order water streams (i.e., formed by study region were randomly assigned to either a cross- the convergence of two second-order streams, which in turn sectional sample or longitudinal cohort. Cross-sectional are formed by two first-order streams arising de novo), which homesteads were visited once during surveys carried out five previous analyses have shown to be predictive of malaria risk times per calendar year to account for temporal differences in in this region,44 was estimated from a digital elevation model. transmission. Homesteads in the longitudinal cohort were Directly observed house features were recorded, including surveyed every other month five times per calendar year on wall composition (fired or unfired brick, cement block, mud average. The current analysis is restricted to participants re- brick, grass, mud, and wooden pole), roof material (iron sheet siding in homesteads enrolled in the cross-sectional survey or corrugated tin, grass, thatch, asbestos sheets), and floor and to the first study visit of participants in homesteads en- (cement, dirt, vinyl, other). Each participant was assessed for rolled in the longitudinal cohort (Figure 1). The study was ap- Plasmodium falciparum infection by rapid diagnostic test proved by the Tropical Diseases Research Center Ethics (RDT) (ICT Diagnostics, Cape Town, South Africa). Individuals Review Committee and the Institutional Review Board at the who tested positive were offered treatment with artemether– Johns Hopkins Bloomberg School of Public Health. Approvals lumefantrine (Coartem®) per World Health Organization and were also obtained from community leaders. national guidelines.3 Homesteads with multiple contempora- Data collection. Quickbird™ satellite images acquired from neous RDT-positive individuals were delineated to account for DigitalGlobe Services, Inc. (Denver, CO) were used to construct clustering of cases at the homestead level.

FIGURE 1. Study ﬂow diagram of participant recruitment from cross-sectional and longitudinal surveys in Southern Province, Zambia from 2008 to 2012, showing the proportion of participants with a positive P. falciparum rapid diagnostic test (RDT). HOUSE STRUCTURE AND MALARIA IN SOUTHERN ZAMBIA 1563

FIGURE 2. Representative photographs of house types in the study area.

Outcome and exposure. The primary outcome was malaria Occupants of mud and grass houses were generally younger infection in individual house occupants, defined as a positive in age and their houses were less likely to have received IRS, RDT result. A house typology scheme was developed accord- have electricity, or have a head of household with greater than ing to wall and roof construction materials. House types in sixth grade education compared with residents of brick or which £ 3% of the total study population resided were ex- cement block houses (Table 1, Figure 2). ITN use among those cluded from the analysis due to insufficient statistical power in low-quality houses was less common, although ITN use was to examine associations between those house types and not significantly associated with malaria in our sample (ad- malaria prevalence (Supplemental Table 1). Houses were justed odds ratio [OR]: 0.60, 95% confidence interval [CI]: assigned to one of three groups: fired brick or cement block 0.30–1.20, P = 0.15). Higher quality houses had a slightly walls with metal roof (high quality), fired brick or cement block smaller proportion of male occupants compared with the walls with grass roof (medium quality), or mud or grass walls lowest quality houses. Distance to third-order streams was with grass roof (low quality). similar among the different house types. IRS coverage was Statistical analysis. Statistical comparisons of baseline low, with 6.8% of high-, 2.6% of medium-, and none of characteristics across house types and malaria prevalence across the low-quality houses reporting ever having their house seasons were done using one-way analysis of variance or χ2 tests. sprayed. Generalized estimating equations logistic regression models House construction. Most participants (65%) resided in clustered by house were fitted to the data, adjusted for age, gen- houses constructed of brick or cement block walls with grass der, bed net use, prior indoor residual spraying, transmission roofs. Thirty percent lived in brick or cement block houses with season, distance to a third-order stream, education level of the metal roofs, and 4% lived in houses of mud or grass walls and household head, and presence of other individuals in the home- grass roofs. Nearly all (96%) houses lacked electricity and stead with a contemporaneously positive RDT. Collinearity was plumbing. All of the low-quality houses and almost all (91%) determined by evaluation of the variance inflation factor, with val- ofthemedium-qualityhouseshaddirtfloors, compared with ues > 10 interpreted as evidence of collinearity. Statistical analyses 48% of high-quality houses. Over the study period 2008–2012, were conducted using Stata 14.0 (StataCorp, College Station, TX). the proportion of participants residing in high-quality houses increased from 11% to 39%, and the percentage of those RESULTS living in low-quality houses decreased from 6% to 2%. Malaria prevalence. A total of 52 RDT-positive individuals Study participants. The study sample consisted of 2,788 (1.9% of the sample) were identified among 41 of the 866 participants residing in 866 houses among 488 homesteads. houses in 36 of the 488 homesteads. RDT positivity ranged 1564 IPPOLITO AND OTHERS

Seven of the 36 homesteads had multiple RDT-positive residents: one homestead had six positive individuals, another homestead had five, two homesteads had three, and three homesteads had two. In the single homestead with six RDT- positive participants, five of the six resided in the same mud- and-grass house. Within the other homesteads, 13 of the 18 houses were medium quality (mud or grass walls and metal roof), and the remaining five houses were high quality (brick or cement block walls and metal roof). Association between housing quality and malaria prevalence. Compared with low-quality houses constructed of mud or grass walls with grass roofs, residing in a medium- or high-quality house was associated with significantly reduced odds of malaria (OR: 0.26, 95% CI: 0.09–0.73, P = 0.01 for houses with brick/cement block walls with metal roofs;OR:0.22,95%CI:0.09–0.52, P <0.01forhouses with brick/cement block walls with grass roofs) (Table 2, Figure 3). No difference was observed between houses with the same wall type (brick/cement block) but different roof type (metal or grass), despite the presumed presence of open eaves in houses with grass roofs (OR: 1.2, 95% CI: 0.58–2.61, P = 0.58). There was a paucity of houses with mud or grass walls and FIGURE 3. Prevalence of P. falciparum infection in 2,788 partici- metal roofs in the sample (< 1% of the total), precluding sta- pants residing among three different house types in the southern tistical testing for effect measure modification between wall Zambia study site from 2008 to 2012, as determined by rapid diagnostic test (RDT). Error bars represent 95% confidence intervals and roof type. Adjusted models with wall and roof type as estimated from the adjusted model. separate variables showed a significant reduction in the odds of malaria prevalence between wall types (OR: 0.22, 95% CI: from 1.4% among participants residing in high-quality houses 0.09–0.52, P < 0.01 for cement/brick versus mud) but not roof to 8.8% among those in low-quality houses. Malaria preva- types (OR: 1.21, 95% CI: 0.57–2.55, P = 0.62). Floor type was lence declined significantly throughout the study period from not significantly associated with RDT positivity and displayed 7% in 2008 to 4% in 2009, and < 1% each subsequent collinearity with house type, hence it was omitted from the transmission season from 2010 to 2012 (P < 0.001). adjusted model.

TABLE 1 Sociodemographic and household characteristics of the study sample House type

Brick or cement walls and metal roof Brick or cement walls and grass roof Mud or grass walls and grass roof

Characteristic n = 228 n = 593 n =45 P value No. participants (%) 849 (30.5) 1,825 (65.5) 114 (4.1) – Positive RDT, n (%) 12 (1.4) 30 (1.6) 10 (8.8) < 0.01* Age, years, mean (SD) 23.0 (21.5) 21.0 (19.5) 17.0 (14.1) < 0.01* Children £ 5 years old, n (%) 153 (18.0) 392 (21.5) 26 (22.8) 0.10* Male gender, n (%) 375 (44.2) 899 (49.3) 56 (49.1) 0.05* ITN use, n (%)† 255 (30.0) 622 (34.0) 30 (26.3) 0.04* Transmission season, n (%) < 0.01‡ February 2008 to July 2008 22 (11.3) 161 (82.6) 12 (6.2) – August 2008 to July 2009 166 (27.4) 393 (65.0) 46 (7.6) – August 2009 to Jul 2010 267 (31.6) 548 (64.9) 29 (3.4) – August 2010 to July 2011 254 (32.4) 512 (65.3) 18 (2.3) – August 2011 to February 2012 140 (38.9) 211 (58.6) 9 (2.5) – Head of household with > sixth grade 582 (68.6) 1,221 (66.9) 65 (57.0) 0.05* education, n (%) Distance to category 3 stream, meters, 4,595 (2,550) 4,560 (2,270) 4,960 (2,100) 0.20* mean (SD) Floor type, n (%) < 0.01‡ Cement 444 (52.3) 159 (8.7) 0 (0.0) – Dirt 405 (47.7) 1,665 (91.3) 114 (100.0) – House with prior IRS, n (%) 58 (6.8) 47 (2.6) 0 (0.0) < 0.01* House with electricity, n (%) 32 (3.8) 1 (0.1) 0 (0.0) < 0.01* House with ﬂush toilet, n (%) 1 (0.1) 0 (0.0) 0 (0.0) 0.32* IRS = indoor residual spraying; ITN = insecticide-treated bed net; RDT = rapid diagnostic test for P. falciparum; SD = standard deviation. The number below each house type refers to the number of houses among the 488 homesteads within the southern Zambia study site, surveyed between 2008 and 2012. * P value was computed by one-way analysis of variance. † Determined by afﬁrmative answer to the survey item, “Do you sleep under a bed net?” ‡ P value was computed by χ2 test. HOUSE STRUCTURE AND MALARIA IN SOUTHERN ZAMBIA 1565

TABLE 2 Adjusted associations between house construction and malaria prevalence by rapid diagnostic test in the southern Zambia study site from 2008 to 2012 Wall material Roof material No. of participants (% total) Malaria prevalence, % (n) Adjusted OR 95% CI P value Fired brick/cement Iron sheet/corrugated tin 849 (30) 1.4 (12) 0.26 0.09–0.73 0.01 Fired brick/cement Grass 1,825 (65) 1.6 (30) 0.22 0.09–0.52 < 0.01 Mud/grass Grass 114 (4) 8.8 (10) REF CI = conﬁdence interval; OR = odds ratio. Generalized estimating equations logistic regression model adjusted for transmission season, clustering by house, distance to third-order stream, participant age, participant gender, participant bed net use, prior indoor residual spraying, head of household level of education, and homestead with > 1 contemporaneous case.

DISCUSSION toward significance, but appeared underpowered to con- clusively examine roof type due to low numbers of grass or This observational study of the association between malaria thatch roofs among the sample (4% and 12% of sampled prevalence and house structure showed that in a low- houses).15,35 Grass roofs on houses in Macha tend to be transmission area of Southern Province, Zambia, better hous- densely packed and hence may pose an effective barrier to ing was associated with reduced odds of P. falciparum infection entry, whereas walls made of mud or grass may have gaps (defined as a positive RDT). Housing improvements may offer an permitting mosquito ingress. It may therefore be that gaps in effective addition to current vector control approaches. walls associated with mud and grass construction are a Malaria prevalence among occupants of mud-and-grass greater determinant of mosquito vector entry than grass roofs houses was 8.8% compared with 1.4–1.6% among those in the study site. living in houses with brick or cement block walls and metal or The strengths of the study include its large sample size, grass roofs. Controlling for clustering by house, transmission random selection from a satellite-imaged sampling frame, season, and individual- and house-level variables of age, adjustments for several confounders, and generalizability to gender, ITN use, proximity to third-order streams, prior IRS similar low-transmission areas in sub-Saharan Africa. There application, level of education, and cohabitation with an RDT- are also limitations. This was an observational study, limiting positive person, residing in a house of cement or brick walls causal inferences between housing and malaria. The study and metal or grass roof was associated with a significant re- site did not have a sufficient number of houses with the duction in the odds of malaria by approximately 75% com- combination of mud or grass walls and metal roof, limiting pared with residing in a house with mud or grass walls and a the ability to isolate the effect of roof type. Interpretation of grass roof. These results are similar to previous reports of the model for wall and roof type as distinct variables was limited housing and malaria in sub-Saharan Africa, which ranged by collinearity because nearly all houses with high-quality roofs from 15% to 63% risk or odds reduction for high-quality wall had high-quality walls (Supplemental Table 1). Wealth indi- and roof types.29–36 The magnitude of the association ap- cators were measured but the overall level of poverty in the pears comparable to IRS, for which a Cochrane review es- sample precluded additional analyses of household wealth, timated a 74% reduction in parasite prevalence, and ITNs, house structure, and malaria. There were no data on eaves, estimated by a Cochrane review to confer a protective effi- although houses in the study area with grass roofs typically cacy of 13%.45,46 have open eaves while most houses with metal roofs have The observed reduction in odds of malaria infection may be closed or partially blocked eaves. Nor were there data on other attributable to fewer gaps in houses constructed of brick or potentially influential house features such as windows, num- cement blocks and absence of eaves in houses with metal ber of rooms, or ceiling. roofs, leading to reduced mosquito entry.47–49 Previous At the end of the nineteenth century, the Italian malariologist studies in similar settings have shown higher quality walls and Angelo Celli recognized malaria as a disease of rural poverty roofs to be associated with lower indoor mosquito numbers and conducted the first interventional trial against malaria,9 compared with houses with gaps in the walls and grass demonstrating the effectiveness of house modifications that roofs.15,50 Grass roofs have also been associated with longer reduced mosquito entry. With the advent of chemical insec- mosquito survival compared with other roof types,51 and may ticides in the first half of the last century, interest in the basic promote indoor resting and parasite development in the outfitting of houses with screens and structural improvements mosquito, perhaps due to lower temperatures inside houses waned.54 Today, insecticide resistance threatens malaria with grass roofs compared with metal roofs.52 Different house control and elimination.4 Interventions not reliant on insecti- types may encourage or discourage ITN use by promoting or cides, such as housing improvements, could aid in sustaining impeding ventilation throughout the house via windows and progress toward malaria control and elimination.6,11,13,36 Al- open eaves,53 although the current analysis showed ITN use though governments and aid organizations cannot wholesale to be lowest in houses expected to have greater ventilation raise the socioeconomic status of people living in malarious (i.e., low-quality houses). areas, they can nonetheless opt to direct resources toward These results suggest that wall construction may have a combatting malaria in a manner that also elevates standards greater influence on malaria risk than roof construction, al- of living, offering the downstream health, social, and eco- though the low number of house types consisting of low- nomic benefits accompanying that rise. The findings of this quality walls and high-quality roofs limited the analysis. Some study support housing improvements as a worthwhile con- but not all previous studies of housing in sub-Saharan Africa sideration for malaria control efforts, particularly in the face of have similarly found a protective effect of brick walls com- emerging insecticide resistance, and corroborate prior stud- pared with mud walls, but not of metal roofs compared with ies’ findings of a protective benefit of comparable magnitude grass or thatch roofs.15,19,30,35 Two studies showed a trend to current vector control strategies. 1566 IPPOLITO AND OTHERS

Received April 11, 2017. Accepted for publication May 31, 2017. house screening interventions on exposure to malaria vectors and on anaemia in children in The Gambia: a randomised Published online August 7, 2017. controlled trial. Lancet 374: 998–1009. Note: Supplemental table appears at www.ajtmh.org. 14. Kampango A, Braganca M, Sousa B, Charlwood JD, 2013. Net- ting barriers to prevent mosquito entry into houses in southern Acknowledgments: We thank members of the community for their Mozambique: a pilot study. Malar J 12: 99. volunteer participation in the survey and the Macha Research Trust fi 15. Adiamah JH, Koram KA, Thomson MC, Lindsay SW, Todd J, eld team for conducting the survey, without whom this research Greenwood BM, 1993. Entomological risk factors for severe would not be possible. malaria in a peri-urban area of The Gambia. Ann Trop Med Financial support: This work was supported by the Johns Hopkins Parasitol 87: 491–500. Malaria Research Institute, the Bloomberg Family Foundation, and the 16. Koram KA, Bennett S, Adiamah JH, Greenwood BM, 1995. Socio- Division of Microbiology and Infectious Diseases, National Institute of economic determinants are not major risk factors for severe Allergy and Infectious Diseases as part of the International Centers of malaria in Gambian children. Trans R Soc Trop Med Hyg 89: Excellence for Malaria Research (U19AI089680). Matthew M. Ippolito 151–154. was additionally supported by the National Institute of General Med- 17. Brooker S, Clarke S, Njagi JK, Polack S, Mugo B, Estambale B, ical Sciences (T32GM066691) and the Sherrilyn and Ken Fisher Center Muchiri E, Magnussen P, Cox J, 2004. Spatial clustering of for Environmental Infectious Diseases, Division of Infectious Diseases malaria and associated risk factors during an epidemic in a of the Johns Hopkins University School of Medicine. highland area of western Kenya. Trop Med Int Health 9: – ’ 757 766. Authors addresses: Matthew M. Ippolito, Divisions of Infectious 18. Nkuo-Akenji T, Ntonifor NN, Ndukum MB, Abongwa EL, Diseases and Clinical Pharmacology, Department of Medicine, Johns Nkwescheu A, Anong DN, Songmbe M, Boyo MG, Ndamukong Hopkins University School of Medicine, Baltimore, MD, E-mail: KN, Titanji VP, 2006. Environmental factors affecting malaria [email protected]. Kelly M. Searle, Timothy M. Shields, and William J. parasite prevalence in rural Bolifamba, South West Cameroon. Moss, Department of Epidemiology, Johns Hopkins Bloomberg Afr J Health Sci 13: 40–46. School of Public Health, Baltimore, MD, E-mails: [email protected], 19. Somi MF, Butler JR, Vahid F, Njau J, Kachur SP, Abdulla S, 2007. [email protected], and [email protected]. Harry Hamapumbu and Is there evidence for dual causation between malaria and so- Philip E. Thuma, Macha Research Trust, Macha Hospital, Choma cioeconomic status? Findings from rural Tanzania. Am J Trop District, Zambia, E-mails: [email protected] Med Hyg 77: 1020–1027. and [email protected]. Jennifer C. Stevenson, The 20. Ernst KC, Lindblade KA, Koech D, Sumba PO, Kuwuor DO, John W. Harry Feinstone Department of Molecular Microbiology and Im- CC, Wilson ML, 2009. Environmental, socio-demographic and munology, Johns Hopkins Malaria Research Institute, Johns Hopkins behavioural determinants of malaria risk in the western Kenyan Bloomberg School of Public Health, Baltimore, MD, and Macha Re- highlands: a case-control study. Trop Med Int Health 14: search Trust, Macha Hospital, Choma District, Zambia, E-mail: jennyc. 1258–1265. [email protected]. 21. Siri JG, Wilson ML, Murray S, Rosen DH, Vulule JM, Slutsker L, Lindblade KA, 2010. Significance of travel to rural areas as a risk REFERENCES factor for malarial anemia in an urban setting. Am J Trop Med Hyg 82: 391–397. 1. Bhatt S, et al., 2015. The effect of malaria control on Plasmodium 22. Yamamoto S, Louis VR, Sie A, Sauerborn R, 2010. Household risk falciparum in Africa between 2000 and 2015. Nature 526: 207–211. factors for clinical malaria in a semi-urban area of Burkina Faso: – 2. World Health Organzation, 2015. World Malaria Report. Geneva, a case-control study. Trans R Soc Trop Med Hyg 104: 61 65. Switzerland: WHO Global Malaria Programme. 23. Winskill P, Rowland M, Mtove G, Malima RC, Kirby MJ, 2011. 3. Reyburn H, 2010. New WHO guidelines for the treatment of Malaria risk factors in north-east Tanzania. Malar J 10: 98. malaria. BMJ 340: c2637. 24. Yukich JO, Taylor C, Eisele TP, Reithinger R, Nauhassenay H, 4. Quinones ML, et al., 2015. Insecticide resistance in areas under Berhane Y, Keating J, 2013. Travel history and malaria infection investigation by the International Centers of Excellence for risk in a low-transmission setting in Ethiopia: a case control Malaria Research: a challenge for malaria control and elimina- study. Malar J 12: 33. tion. Am J Trop Med Hyg 93: 69–78. 25. Magalhaes RJ, Langa A, Sousa-Figueiredo JC, Clements AC, 5. Huho B, et al., 2013. Consistently high estimates for the pro- Nery SV, 2012. Finding malaria hot-spots in northern Angola: portion of human exposure to malaria vector populations oc- the role of individual, household and environmental factors curring indoors in rural Africa. Int J Epidemiol 42: 235–247. within a meso-endemic area. Malar J 11: 385. 6. Tusting LS, Ippolito MM, Willey BA, Kleinschmidt I, Dorsey G, 26. Townes LR, Mwandama D, Mathanga DP, Wilson ML, 2013. Ele- Gosling RD, Lindsay SW, 2015. The evidence for improving vated dry-season malaria prevalence associated with fine- housing to reduce malaria: a systematic review and meta- scale spatial patterns of environmental risk: a case-control analysis. Malar J 14: 209. study of children in rural Malawi. Malar J 12: 407. 7. Tusting LS, Bottomley C, Gibson H, Kleinschmidt I, Tatem AJ, 27. Tilaye T, Deressa W, 2007. Prevalence of urban malaria and as- Lindsay SW, Gething PW, 2017. Housing improvements and sociated factors in Gondar Town, Northwest Ethiopia. Ethiop malaria risk in sub-Saharan Africa: a multi-country analysis of Med J 45: 151–158. survey data. PLoS Med 14: e1002234. 28. Ong’echa JM, et al., 2006. Parasitemia, anemia, and malarial 8. Lane C, 1931. Housing and Malaria: A Critical Summary of the anemia in infants and young children in a rural holoendemic Literature Dealing with This Subject. Geneva, Switzerland: Plasmodium falciparum transmission area. Am J Trop Med Hyg League of Nations: Health Organisation. 74: 376–385. 9. Celli A, 1900. The new prophylaxis against malaria in Lazio. Lancet 29. Sintasath DM, Ghebremeskel T, Lynch M, Kleinau E, Bretas G, 156: 1603–1606. Shililu J, Brantly E, Graves PM, Beier JC, 2005. Malaria preva- 10. Harrison G, 1978. Mosquitoes, Malaria, and Man: A History of the lence and associated risk factors in Eritrea. Am J Trop Med Hyg Hostilities Since 1880. Dutton. 72: 682–687. 11. Ogoma SB, Lweitoijera DW, Ngonyani H, Furer B, Russell TL, 30. Bousema T, et al., 2010. Identification of hot spots of malaria Mukabana WR, Killeen GF, Moore SJ, 2010. Screening mos- transmission for targeted malaria control. J Infect Dis 201: quito house entry points as a potential method for integrated 1764–1774. control of endophagic filariasis, arbovirus and malaria vectors. 31. Ouma P, van Eijk AM, Hamel MJ, Parise M, Ayisi JG, Otieno K, PLoS Negl Trop Dis 4: e773. Kager PA, Slutsker L, 2007. Malaria and anaemia among 12. Massebo F, Lindtjorn B, 2013. The effect of screening doors and pregnant women at first antenatal clinic visit in Kisumu, western windows on indoor density of Anopheles arabiensis in south- Kenya. Trop Med Int Health 12: 1515–1523. west Ethiopia: a randomized trial. Malar J 12: 319. 32. Mmbando BP, Kamugisha ML, Lusingu JP, Francis F, 13. Kirby MJ, Ameh D, Bottomley C, Green C, Jawara M, Milligan PJ, Ishengoma DS, Theander TG, Lemnge MM, Scheike TH, Snell PC, Conway DJ, Lindsay SW, 2009. Effect of two different 2011. Spatial variation and socio-economic determinants of HOUSE STRUCTURE AND MALARIA IN SOUTHERN ZAMBIA 1567

Plasmodium falciparum infection in northeastern Tanzania. and prospects for malaria elimination in the southern Africa Malar J 10: 145. region. Acta Trop 121: 207–211. 33. Ye Y, Hoshen M, Louis V, Seraphin S, Traore I, Sauerborn R, 2006. 44. Moss WJ, Hamapumbu H, Kobayashi T, Shields T, Kamanga A, Housing conditions and Plasmodium falciparum infection: Clennon J, Mharakurwa S, Thuma PE, Glass G, 2011. Use of protective effect of iron-sheet roofed houses. Malar J 5: 8. remote sensing to identify spatial risk factors for malaria in a 34. Temu EA, Coleman M, Abilio AP, Kleinschmidt I, 2012. High region of declining transmission: a cross-sectional and longi- prevalence of malaria in Zambezia, Mozambique: the pro- tudinal community survey. Malar J 10: 163. tective effect of IRS versus increased risks due to pig-keeping 45. Lengeler C, 2004. Insecticide-treated bed nets and curtains for and house construction. PLoS One 7: e31409. preventing malaria. Cochrane Database Syst Rev CD000363. 35. Rulisa S, et al., 2013. Malaria prevalence, spatial clustering and 46. Pluess B, Tanser FC, Lengeler C, Sharp BL, 2010. Indoor residual risk factors in a low endemic area of eastern Rwanda: a cross spraying for preventing malaria. Cochrane Database Syst Rev sectional study. PLoS One 8: e69443. CD006657. 36. Wanzirah H, et al., 2015. Mind the gap: house structure and the 47. Kirby MJ, West P, Green C, Jasseh M, Lindsay SW, 2008. Risk risk of malaria in Uganda. PLoS One 10: e0117396. factors for house-entry by culicine mosquitoes in a rural town 37. Ghebreyesus TA, Haile M, Witten KH, Getachew A, Yohannes M, and satellite villages in The Gambia. Parasit Vectors 1: 41. Lindsay SW, Byass P, 2000. Household risk factors for malaria 48. Lindsay SW, Snow RW, 1988. The trouble with eaves; house among children in the Ethiopian highlands. Trans R Soc Trop entry by vectors of malaria. Trans R Soc Trop Med Hyg 82: Med Hyg 94: 17–21. 645–646. 38. Snyman K, et al., 2015. Poor housing construction associated 49. Njie M, Dilger E, Lindsay SW, Kirby MJ, 2009. Importance of eaves with increased malaria incidence in a cohort of young Ugandan to house entry by anopheline, but not culicine, mosquitoes. children. Am J Trop Med Hyg 92: 1207–1213. J Med Entomol 46: 505–510. 39. Sutcliffe CG, Kobayashi T, Hamapumbu H, Shields T, Mharakurwa 50. Animut A, Balkew M, Lindtjorn B, 2013. Impact of housing S, Thuma PE, Louis TA, Glass G, Moss WJ, 2012. Reduced risk condition on indoor-biting and indoor-resting Anopheles of malaria parasitemia following household screening and arabiensis density in a highland area, central Ethiopia. Malar J treatment: a cross-sectional and longitudinal cohort study. PLoS 12: 393. One 7: e31396. 51. Magesa SM, Wilkes TJ, Mnzava AE, Njunwa KJ, Myamba J, Kivuyo 40. Moss WJ, et al., 2015. Malaria epidemiology and control within the MD,HillN,LinesJD,CurtisCF,1991.Trialofpyrethroidimpreg- International Centers of Excellence for Malaria Research. Am J nated bednets in an area of Tanzania holoendemic for malaria. Trop Med Hyg 93: 5–15. Part 2. Effects on the malaria vector population. Acta Trop 49: 41. Chanda E, Kamuliwo M, Steketee RW, Macdonald MB, Babaniyi 97–108. O, Mukonka VM, 2013. An overview of the malaria control 52. Murdock CC, Sternberg ED, Thomas MB, 2016. Malaria trans- programme in Zambia. ISRN Prev Med 2013: 495037. mission potential could be reduced with current and future 42. Chizema-Kawesha E, Miller JM, Steketee RW, Mukonka VM, climate change. Sci Rep 6: 27771. Mukuka C, Mohamed AD, Miti SK, Campbell CC, 2010. Scaling 53. von Seidlein L, Ikonomidis K, Bruun R, Jawara M, Pinder M, Knols up malaria control in Zambia: progress and impact 2005–2008. BG, Knudsen JB, 2012. Airﬂow attenuation and bed net utili- Am J Trop Med Hyg 83: 480–488. zation: observations from Africa and Asia. Malar J 11: 200. 43. Moss WJ, Norris DE, Mharakurwa S, Scott A, Mulenga M, Mason 54. Lindsay SW, Emerson PM, Charlwood JD, 2002. Reducing malaria PR, Chipeta J, Thuma PE, Southern Africa IT, 2012. Challenges by mosquito-prooﬁng houses. Trends Parasitol 18: 510–514.